Safety monitoring method, device and equipment for self-installation offshore oil platform structure and medium

Abstract

The invention relates to the technical field of ocean engineering, and discloses a self-installation offshore oil platform structure safety monitoring method, device and equipment and a medium, and the method comprises the steps: obtaining the preprocessed strain time sequence data of a pile leg in a self-installation offshore oil platform in a first time period; and carrying out stability analysis based on a confidence interval on the preprocessed strain time series data to obtain an analysis result, and identifying and stripping a random load component in the preprocessed strain time series data according to the analysis result to obtain target strain time series data. And performing finite element driven prestress compensation and stress mapping on the target strain time sequence data to obtain key stress time sequence data. And according to the key stress time sequence data, the accumulated fatigue damage index of the pile leg in the second time period is predicted. And performing corrosion coupling correction on the accumulated fatigue damage index to obtain a target damage index, and judging whether to perform fatigue early warning or not according to the target damage index. According to the invention, the reliability of safety monitoring of the self-installation offshore oil platform structure can be enhanced.

Description

Technical Field

[0001] This invention relates to the field of marine engineering technology, and in particular to a method, device, equipment and medium for structural safety monitoring of self-installing offshore oil platforms. Background Technology

[0002] With the development of science and technology, marine engineering technology is constantly improving.

[0003] Self-installing offshore oil platforms are a type of efficient and economical fixed offshore oil platform that achieves rapid installation through self-weight settling and leg fixation, making them suitable for marginal oil fields.

[0004] Self-installed offshore oil platforms operate in complex and ever-changing marine environments, including adverse factors such as tidal currents, wind loads, wave impacts, freezing, seismic vibrations, corrosion, material aging, and collisions with other vessels. The interaction of these adverse factors can lead to a gradual decrease in the platform's structural resistance, potentially resulting in structural failure. Therefore, relevant technologies are needed to conduct real-time and routine safety monitoring of the platform to ensure production safety and the safety of personnel.

[0005] However, the structural characteristics of self-installed offshore oil platforms make safety monitoring difficult and the reliability of safety monitoring low. Summary of the Invention

[0006] This invention provides a method, apparatus, equipment, and medium for structural safety monitoring of self-installed offshore oil platforms, which addresses the shortcomings of related technologies where the structural characteristics of self-installed offshore oil platforms make safety monitoring difficult and unreliable, thereby enhancing the reliability of safety monitoring.

[0007] In a first aspect, the present invention provides a method for structural safety monitoring of a self-installing offshore oil platform, comprising: Acquire pre-processed strain time-series data of the legs of the installed offshore oil platform during the first time period; A stationarity analysis based on confidence intervals is performed on the preprocessed strain time series data to obtain the analysis results. Based on the analysis results, the random load components in the preprocessed strain time series data are identified and removed to obtain the target strain time series data. Finite element-driven prestress compensation and stress mapping are performed on the target strain time series data to obtain key stress time series data; The cumulative fatigue damage index of the pile leg in the second time period is predicted based on the key stress time series data; wherein, the second time period is after the first time period; The cumulative fatigue damage index is subjected to corrosion coupling correction to obtain the target damage index, and fatigue warning is determined based on the target damage index.

[0008] Optionally, the preprocessed strain time series data includes preprocessed strain values ​​corresponding to multiple consecutive timestamps; The step of performing stationarity analysis based on confidence intervals on the preprocessed strain time series data to obtain analysis results includes: Calculate the mean and standard deviation using the preprocessed strain value corresponding to each timestamp; The preprocessed strain value, mean, and standard deviation corresponding to each timestamp are input into the created confidence interval calculation model to determine the confidence interval corresponding to each timestamp; For any of the timestamps, determine whether there is an intersection between the confidence interval corresponding to the timestamp and the confidence interval corresponding to the previous timestamp; If there is an intersection, it is determined that the stability of the preprocessed strain value corresponding to the timestamp meets the requirements; If there is no intersection, it is determined that the stability of the preprocessed strain value corresponding to the timestamp does not meet the requirements.

[0009] Optionally, determining whether the confidence interval corresponding to the timestamp and the confidence interval corresponding to the previous timestamp intersect includes: A first upper bound and a first lower bound are determined in the confidence interval corresponding to the timestamp, and a second upper bound and a second lower bound are determined in the confidence interval corresponding to the previous timestamp. Determine the minimum value between the first upper bound and the second upper bound, and determine the maximum value between the first lower bound and the second lower bound; Subtract the maximum value from the minimum value to obtain the corresponding difference; The maximum value between 0 and the difference is selected as the intersection measure; If the intersection measure is not equal to 0, then it is determined that the confidence interval corresponding to the timestamp and the confidence interval corresponding to the previous timestamp have an intersection. If the intersection measure is equal to 0, then it is determined that the confidence interval corresponding to the timestamp and the confidence interval corresponding to the previous timestamp do not intersect.

[0010] Optionally, the step of identifying and removing the random load components from the preprocessed strain time series data based on the analysis results to obtain the target strain time series data includes: Each pre-processed strain value whose stability requirement is not met in the pre-processed strain time series data is determined as a random load component. Each of the random load components in the preprocessed strain time series data is deleted to obtain the target strain time series data.

[0011] Optionally, the step of performing finite element-driven prestress compensation and stress mapping on the target strain time series data to obtain key stress time series data includes: A three-dimensional model of the pile leg is constructed based on the finite element method and the set model conditions. The self-weight stress of the pile leg is simulated using the three-dimensional model to obtain the self-weight stress distribution of the pile leg. The target strain time series data and the stress distribution of the pile leg self-weight are input into the created mapping model to perform prestress compensation and stress mapping, thereby obtaining the key stress time series data. The mapping model is as follows: ; in, The key stress time series data, To determine through static simulation using finite element modeling, For target strain time series data, The stress distribution of the self-weight of the pile leg is shown. For corrosion reduction factor, To obtain corrosion stress time series data.

[0012] Optionally, predicting the cumulative fatigue damage index of the pile leg in the second time period based on the key stress time series data includes: The key stress time series data are input into the trained time series prediction model so that the time series prediction model can predict the cumulative fatigue damage index of the pile leg in the second time period. The step of performing corrosion coupling correction on the cumulative fatigue damage index to obtain the target damage index includes: The corresponding corrosion coupling index is determined based on the cumulative fatigue damage index. The cumulative fatigue damage index is added to the corrosion coupling index to obtain the corresponding target damage index; The step of determining whether to issue a fatigue warning based on the target damage index includes: Determine whether the target damage index is greater than a set threshold; If the target damage index is greater than the set threshold, a fatigue warning will be issued; If the target damage index is not greater than the set threshold, fatigue warning is prohibited.

[0013] Optionally, the acquisition of pre-processed strain time-series data from the legs installed on the offshore oil platform during the first time period includes: Obtain the original strain time series data of the pile leg during the first time period; The original strain time series data is subjected to outlier identification and removal to obtain strain time series data after removal; Based on the moving average window, the strain time series data after removal is smoothed and filtered to obtain the filtered strain time series data, which is used as the preprocessed strain time series data.

[0014] Secondly, the present invention provides a self-installing offshore oil platform structural safety monitoring device, comprising: The acquisition unit is used to acquire the pre-processed strain time series data of the legs of the self-installed offshore oil platform during the first time period. The analysis unit is used to perform stationarity analysis based on confidence intervals on the preprocessed strain time series data to obtain analysis results; The stripping unit is used to identify and strip the random load components in the preprocessed strain time series data according to the analysis results, so as to obtain the target strain time series data. The mapping unit is used to perform finite element-driven prestress compensation and stress mapping on the target strain time series data to obtain key stress time series data. The prediction unit is used to predict the cumulative fatigue damage index of the pile leg in a second time period based on the key stress time series data; wherein the second time period is after the first time period; A correction unit is used to perform corrosion coupling correction on the cumulative fatigue damage index to obtain the target damage index; The early warning unit is used to determine whether to issue a fatigue warning based on the target damage index.

[0015] Thirdly, the present invention provides a computer device, comprising: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the self-installing offshore oil platform structure safety monitoring method described in the first aspect or any corresponding embodiment.

[0016] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to execute the self-installing offshore oil platform structural safety monitoring method described in the first aspect or any corresponding embodiment thereof.

[0017] The present invention provides a method, apparatus, equipment, and medium for structural safety monitoring of self-installed offshore oil platforms. This method can acquire pre-processed strain time-series data of the legs in a self-installed offshore oil platform during a first time period. It performs stationarity analysis based on confidence intervals on the pre-processed strain time-series data to obtain analysis results. Based on the analysis results, it identifies and removes random load components from the pre-processed strain time-series data to obtain target strain time-series data. Finite element-driven prestress compensation and stress mapping are then performed on the target strain time-series data to obtain key stress time-series data. The cumulative fatigue damage index of the legs during a second time period is predicted based on the key stress time-series data; the second time period is after the first time period. Corrosion coupling correction is applied to the cumulative fatigue damage index to obtain the target damage index, and a fatigue warning is determined based on the target damage index. This invention achieves integrated confidence interval analysis, prestress compensation mapping, and fatigue damage prediction for safety monitoring, thereby realizing high-precision, low-power real-time monitoring and enhancing the reliability of structural safety monitoring for self-installed offshore oil platforms. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in this invention or related technologies, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 A flowchart illustrating a method for monitoring the structural safety of a self-installing offshore oil platform, as provided in an embodiment of the present invention; Figure 2 This is a comparison chart of the characteristics of different monitoring methods provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of a self-installing offshore oil platform structural safety monitoring system provided in an embodiment of the present invention; Figure 4 A schematic diagram of digital twin three-dimensional visualization provided in an embodiment of the present invention; Figure 5 This is a schematic diagram of a self-installing offshore oil platform structural safety monitoring device provided in an embodiment of the present invention; Figure 6 This is a schematic diagram of the structure of a computer device provided in an embodiment of the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0021] The following is combined with Figures 1-4 This invention describes a method for monitoring the structural safety of self-installing offshore oil platforms.

[0022] In related technologies, structural health monitoring using self-installed platforms has gone through three stages of development: the first stage was independent instrument monitoring, using large instruments to monitor equipment parameters in real time, but this consumed a lot of manpower and resources and had low integration; the second stage was a centralized system with physical connections, monitoring subsystems through display devices, but the independent logical functions were not conducive to fault location; the third stage is a shift to intelligent integrated platforms, utilizing the Internet of Things, edge computing, and cloud platforms to achieve full network monitoring, supporting multi-dimensional data analysis and fault prediction. According to relevant reports, current structural health monitoring systems mainly rely on sensor networks (such as strain gauges, inclinometers, and accelerometers) to collect vibration, stress, and displacement data, and use threshold methods or spectral analysis to determine anomalies.

[0023] However, the relevant technologies have significant limitations: (1) Problem of random load interference: Dynamic loads such as wind, waves and currents cause high noise in settlement monitoring data. The false alarm rate of traditional threshold method is as high as 20%-30%, which cannot effectively remove environmental noise. Literature review shows that although the monitoring accuracy of Global Navigation Satellite System (GNSS) reaches the millimeter level, it lacks statistical methods such as confidence intervals to quantify uncertainty.

[0024] (2) Insufficient prestressing treatment: The strain gauges on the pile legs only measure the stress change and ignore the initial prestress, resulting in an absolute stress assessment deviation of >10%. Related technologies focus on vibration / settlement, but do not consider the prestress benchmark calibration of the self-installation platform, especially when the pile legs are already in place (project risk analysis). Related technologies emphasize that the pile leg design should consider prestress, but the monitoring methods mostly rely on manual adjustment, which is susceptible to corrosion.

[0025] (3) Lack of intelligent early warning: Traditional systems rely on rule-based judgments, ignoring the time-based accumulation of fatigue damage and corrosion coupling. Related technologies require SN curves to assess fatigue life, but their application is superficial and lacks transfer learning and big data early warning for self-installation platforms. Related technologies use computer vision, but do not integrate prestress compensation; they assess anomalies but do not have confidence interval stripping. Related technologies focus on installation loads, but there is little innovation in structural health monitoring, and statistical analysis and application are significantly lacking.

[0026] (4) Poor system integration and environmental adaptability: The existing system has high power consumption (>500W) and is not suitable for operation during typhoon season; the protection level is insufficient and it is easily affected by high temperature and high salt spray.

[0027] In summary, there is an urgent need for a structural health monitoring method that integrates confidence interval analysis (removing random effects), prestress compensation (absolute stress mapping), and fatigue damage prediction to fill the technological gap and support the intelligent operation and maintenance of self-installation platforms.

[0028] like Figure 1 As shown in the figure, this embodiment proposes a first method for structural safety monitoring of self-installing offshore oil platforms, which may include the following steps: S101. Obtain the pre-processed strain time series data of the legs of the self-installed offshore oil platform during the first time period.

[0029] The first time period can be a time period designated by the technical personnel.

[0030] Specifically, the preprocessed strain time series data may include preprocessed strain values ​​from multiple consecutive timestamps within the first time period. The preprocessed strain values ​​are obtained by preprocessing the original strain values.

[0031] Optionally, step S101 includes: Obtain the original strain time series data of the pile leg during the first time period; Outlier identification and removal are performed on the original strain time series data to obtain strain time series data after removal. Based on the moving average window, the strain time series data after rejection is smoothed and filtered to obtain the filtered strain time series data, which is used as the preprocessed strain time series data.

[0032] Specifically, the raw strain time series data includes raw strain values ​​from multiple consecutive timestamps within the first time period.

[0033] In this embodiment, the original strain value of the pile leg can be collected by a sensor installed on the pile leg.

[0034] Specifically, this embodiment can use 3 The criteria identify and remove outliers from the original strain time series data.

[0035] In this embodiment, a moving average window can be used for smoothing filtering. Specifically, in this embodiment, each strain value in the strain time series data after filtering can be input into the following formula: ; The window can be set to 5 minutes. N It can be set to 300 points. t Here is the current timestamp (1 second resolution), and i is the index within the window. This is the filtered strain value obtained by smoothing a strain value in the strain time series data after removing strains. N Based on the Nyquist sampling theorem and the number of data points after preprocessing at a sampling rate of 10Hz, no information loss is ensured. Wind, wave, and current data (obtained from a weather station) are used as auxiliary input for subsequent stripping. This filtering reduces noise by 20dB.

[0036] S102. Perform stationarity analysis based on confidence intervals on the preprocessed strain time series data to obtain the analysis results.

[0037] Specifically, this embodiment can perform stationarity analysis based on confidence intervals on each pre-processed strain value in the pre-processed strain time series data to obtain the analysis results corresponding to each pre-processed strain value.

[0038] Optionally, the preprocessed strain time series data includes preprocessed strain values ​​corresponding to multiple consecutive timestamps. Step S102 includes: Calculate the mean and standard deviation using the preprocessed strain value corresponding to each timestamp; The preprocessed strain value, mean, and standard deviation corresponding to each time stamp are input into the created confidence interval calculation model to determine the confidence interval corresponding to each time stamp; For any timestamp, determine whether there is an intersection between the confidence interval corresponding to the timestamp and the confidence interval corresponding to the previous timestamp; If there is an intersection, it is determined that the stationarity of the preprocessed strain value corresponding to the timestamp meets the requirements; If there is no intersection, it is determined that the stationarity of the preprocessed strain value corresponding to the timestamp does not meet the requirements.

[0039] In this embodiment, the mean can be calculated using the following formula. and standard deviation : .

[0040] The confidence interval calculation model (95% confidence level, z=1.96) is as follows: .

[0041] in, For a given timestamp, the confidence interval is... This represents the preprocessed strain value corresponding to a specific timestamp. The sample size for the time window is 100 points.

[0042] Optionally, the above determination of whether the confidence interval corresponding to the timestamp and the confidence interval corresponding to the previous timestamp intersect includes: Determine the first upper bound and the first lower bound in the confidence interval corresponding to the timestamp, and determine the second upper bound and the second lower bound in the confidence interval corresponding to the previous timestamp; Determine the minimum value between the first upper bound and the second upper bound, and determine the maximum value between the first lower bound and the second lower bound; Subtract the maximum value from the minimum value to obtain the corresponding difference; Select the maximum value between 0 and the difference and use it as the intersection measure; If the intersection measure is not equal to 0, then it is determined that the confidence interval corresponding to the timestamp and the confidence interval corresponding to the previous timestamp have an intersection. If the intersection measure is equal to 0, it is determined that the confidence interval corresponding to the timestamp and the confidence interval corresponding to the previous timestamp do not intersect.

[0043] Specifically, this embodiment can determine whether there is an intersection between confidence intervals corresponding to adjacent timestamps using a formula. The formula is: ; in, This is a measure of the intersection of adjacent timestamps. and These are the upper bounds of the confidence intervals corresponding to the timestamp and the previous timestamp, respectively. and These are the lower bounds of the confidence intervals corresponding to the timestamp and the previous timestamp, respectively.

[0044] S103. Based on the analysis results, identify and remove the random load components in the preprocessed strain time series data to obtain the target strain time series data.

[0045] Specifically, in this embodiment, the random load component in the pre-processed strain time series data can be identified and removed based on the stationarity analysis results of each pre-processed strain value in the pre-processed strain time series data to obtain the target strain time series data.

[0046] Optionally, step S103 includes: Each pre-processed strain value in the pre-processed strain time series data that does not meet the stationarity requirement is identified as a random load component. Each random load component in the preprocessed strain time series data is deleted to obtain the target strain time series data.

[0047] Specifically, in this embodiment, each pre-processed strain value whose stability requirement is not met in the pre-processed strain time series data can be identified as a random load component. Each identified random load component is deleted from the pre-processed strain time series data, and the remaining ordered parameter values ​​in the pre-processed strain time series data are used as the target strain time series data.

[0048] It should be noted that confidence interval analysis differs from traditional spectral analysis. Confidence interval analysis is statistically robust and applicable to marine stochastic processes.

[0049] In practical applications, this embodiment can collect multi-source time-series data including settlement, strain, vibration, and tilt angle. The multi-source time-series data is preprocessed and stationarity analysis is performed based on confidence intervals. Based on the stationarity analysis results, multi-dimensional safety monitoring of the self-installation platform can be achieved, such as settlement monitoring. When the monitoring result of a certain dimension is abnormal, a corresponding early warning can be triggered in a timely manner, such as a settlement early warning.

[0050] S104. Perform finite element-driven prestress compensation and stress mapping on the target strain time series data to obtain key stress time series data.

[0051] Specifically, in this embodiment, after obtaining the target strain time series data, the key stress time series data can be obtained by performing finite element-driven prestress compensation and stress mapping on the target strain time series data.

[0052] Optionally, step S104 includes: A three-dimensional model of the pile leg is constructed based on the finite element method and the set model conditions. The self-weight stress of the pile leg is simulated using the three-dimensional model to obtain the self-weight stress distribution of the pile leg. The target strain time series data and the stress distribution of the pile leg self-weight are input into the created mapping model to perform prestress compensation and stress mapping, and key stress time series data are obtained. The mapping model is as follows: ; in, For key stress time series data, To determine through static simulation using finite element modeling, For target strain time series data, The stress distribution due to the self-weight of the pile leg. For corrosion reduction factor, To obtain corrosion stress time series data.

[0053] Specifically, this embodiment can perform finite element modeling to construct a 3D model of the pile leg (material API 5L X65 steel). Mesh size 5mm, boundary condition: bottom fixed. Applied load: self-weight + wave (Morison equation, , ).simulation Initial distribution: ; P For axial force, M For bending moment, A and I These are the cross-sectional parameters. Output the hot spot stress field with an accuracy of <1MPa.

[0054] Regarding the mapping relationship, this embodiment can establish the strain at the measurement point (the strain measured by the fiber optic / strain gauge). Key stress Linear / nonlinear mapping: ; k Fitted by finite element modeling, This is a corrosion reduction factor. The compensation error is <5%, which is better than manual calibration.

[0055] S105. Predict the cumulative fatigue damage index of the pile leg in the second time period based on the key stress time series data; wherein, the second time period is after the first time period.

[0056] The second time period is the period following the first time period, such as a period 24 hours apart. The second time period can be specified by the technicians, but this embodiment does not limit it.

[0057] Specifically, this embodiment can use key stress time series data to predict the cumulative fatigue damage index of the pile leg in the second time period.

[0058] Optionally, step S105 includes: Key stress time series data are input into a trained time series prediction model so that the time series prediction model can predict the cumulative fatigue damage index of the pile leg in the second time period.

[0059] Among them, the time-series prediction model can be a Long Short-Term Memory (LSTM) network.

[0060] Specifically, LSTM includes 128 hidden units, dropout=0.2, and also a fully connected layer.

[0061] Specifically, in this embodiment, a training dataset can be generated through historical monitoring and intelligent synthesis. The training dataset is then used to train the time series prediction model, calculate the loss function, and update the LSTM parameters based on the loss function until a well-trained time series prediction model is obtained.

[0062] S106. Corrosion coupling correction is applied to the cumulative fatigue damage index to obtain the target damage index.

[0063] Specifically, step S106 includes: The corresponding corrosion coupling index is determined based on the cumulative fatigue damage index. The cumulative fatigue damage index is added to the corrosion coupling index to obtain the corresponding target damage index.

[0064] Specifically, this embodiment can determine the corresponding corrosion coupling index based on the Paris law of subcritical crack propagation and the cumulative fatigue damage index. The cumulative fatigue damage index and the corrosion coupling index are added together to obtain the corresponding target damage index, thereby realizing corrosion coupling correction and improving the accuracy of fatigue damage index evaluation.

[0065] S107. Determine whether to issue a fatigue warning based on the target damage index.

[0066] Specifically, step S107 includes: Determine whether the target damage index is greater than a set threshold; If the target damage index is greater than the set threshold, a fatigue warning will be issued; If the target damage index is not greater than the set threshold, fatigue warning is prohibited.

[0067] The threshold value can be set by technicians according to the actual situation, such as 0.8, but this embodiment does not limit it.

[0068] The structural safety monitoring method for self-installing offshore oil platforms proposed in this embodiment can acquire pre-processed strain time-series data of the legs in the self-installing offshore oil platform during the first time period. Stationarity analysis based on confidence intervals is performed on the pre-processed strain time-series data to obtain the analysis results. Based on the analysis results, random load components in the pre-processed strain time-series data are identified and removed to obtain target strain time-series data. Finite element-driven prestress compensation and stress mapping are performed on the target strain time-series data to obtain key stress time-series data. The cumulative fatigue damage index of the legs during the second time period is predicted based on the key stress time-series data; the second time period is after the first time period. Corrosion coupling correction is applied to the cumulative fatigue damage index to obtain the target damage index, and fatigue warning is determined based on the target damage index. This embodiment achieves integrated confidence interval analysis, prestress compensation mapping, and fatigue damage prediction for safety monitoring, thereby realizing high-precision, low-power real-time monitoring and enhancing the reliability of structural safety monitoring for self-installing offshore oil platforms.

[0069] based on Figure 1 In other self-installing offshore oil platform structural safety monitoring methods proposed in this embodiment, the method is applied to a safety monitoring system, which includes a sensor module, a data acquisition module, a data processing module, an early warning module, and a digital twin module; wherein: The sensor module includes a variety of sensors for collecting data on the platform's settlement, strain, tilt angle, and vibration. It is made of corrosion-resistant materials, has a high protection level, and is suitable for a variety of marine environments. The data acquisition module transmits data via wireless communication technology and supports integration with existing instruments on the platform; The data processing module is used to perform confidence interval analysis, prestress compensation, and machine learning fatigue prediction to generate structural health status. The early warning module triggers alarms and pushes reports based on confidence interval analysis results or fatigue damage predictions; The digital twin module is used to dynamically update the platform structure model and supports predictive maintenance.

[0070] Optionally, the sensor module includes strain gauges, inclinometers, positioning modules, and vibration sensors, which are respectively installed on the platform legs, deck, and top key parts. It has high sensitivity and redundancy design, and is used to collect multi-dimensional structural response data. It is suitable for high salinity and high pressure marine environments.

[0071] Optionally, the data acquisition module includes a wireless transmission unit and an edge computing unit; the wireless transmission unit adopts an encryption protocol and supports concurrent data transmission across multiple nodes; the edge computing unit performs data preprocessing, including outlier removal and smoothing filtering to reduce the impact of noise.

[0072] Optionally, the data acquisition module can be integrated with the platform's existing instruments via a standard communication protocol, supporting data fusion, with low modification costs, and suitable for various marine engineering scenarios.

[0073] Optionally, the data processing module includes a finite element modeling unit, a confidence interval analysis unit, and a machine learning unit; the finite element modeling unit establishes a platform structure model based on high-strength material parameters to simulate multi-source loads; the confidence interval analysis unit performs statistical analysis on the monitoring data to quantify uncertainty; and the machine learning unit predicts structural fatigue damage.

[0074] Optionally, the confidence interval analysis unit smooths the monitoring data through a sliding window and calculates the confidence interval. When the real-time data does not intersect with the baseline interval, it determines the structural anomaly and removes the influence of random loads.

[0075] Optionally, prestress compensation determines the initial stress field through finite element modeling, establishes a stress mapping relationship by combining sensor measurement data, and controls the calibration error within a safe range, making it suitable for various marine platforms.

[0076] Optionally, the machine learning unit employs a Long Short-Term Memory (LSTM) network, taking time-series stress data and environmental factors as input, and predicts damage based on the fatigue accumulation criterion, supporting early warning and conforming to international marine engineering standards.

[0077] Optionally, the early warning module includes an anomaly detection unit and a report generation unit; the anomaly detection unit triggers alarms based on confidence interval analysis or fatigue damage prediction, with a low false alarm rate; the report generation unit generates visual reports to support on-site and remote decision-making.

[0078] Optionally, the digital twin module includes a model update unit and a visualization unit; the model update unit dynamically updates the structural model based on real-time data; the visualization unit generates a three-dimensional visualization of stress field, settlement trend and fatigue damage, supporting predictive maintenance.

[0079] Optionally, the system adopts a low-power design to reduce energy consumption, accurately monitor and reduce maintenance frequency, minimize marine ecological disturbance, and comply with international environmental protection standards.

[0080] Optionally, the above method may include the following steps: (a) Deploy sensor modules to collect multimodal structural response data and perform preprocessing; (b) Perform confidence interval analysis to assess platform stability; (c) Perform prestress compensation and finite element stress mapping; (d) Predict fatigue damage and trigger early warnings based on machine learning; (e) Integrate long-term monitoring data to generate structural health reports; (f) Update the digital twin model to support predictive maintenance and cross-platform applications.

[0081] Optionally, in step (b), the confidence interval analysis uses a sliding window smoothing process to calculate the confidence interval, judges anomalies based on statistical criteria, and removes the influence of random loads.

[0082] Optionally, in step (c), prestress compensation simulates multi-source loads through finite element modeling and combines sensor data to fit stress mapping relationships, resulting in low calibration error and suitability for various platform materials and structures.

[0083] Optionally, in step (d), the machine learning prediction is based on a long short-term memory network, coupled with a fatigue accumulation criterion, inputting time-series data and environmental factors, and outputting damage prediction, supporting early warning and conforming to international marine engineering standards.

[0084] This embodiment enables structural safety monitoring of self-installing offshore oil platforms based on confidence interval analysis and prestress compensation. It quantifies settlement uncertainty using statistical methods, calibrates pile leg stress through mechanical compensation, and predicts fatigue damage, achieving high-precision, low-power real-time monitoring that supports position monitoring and fatigue assessment requirements. Ultimately, it improves the platform's safety factor by more than 20% and reduces the false alarm rate to <5%, providing technical support for the intelligent operation of offshore oil equipment.

[0085] Specifically, the core of this embodiment lies in confidence interval settlement analysis (stripping random loads from wind, waves, and currents), prestressed dynamic compensation (absolute stress mapping of pile legs), and fatigue early warning (coupling corrosion effects). It is applicable to key structures of self-installing platforms (such as pile legs, water-proof sleeves, and decks), integrates existing instruments, and achieves embedded compatibility.

[0086] Specifically, the system composition is detailed below: The system consists of four modules to ensure a closed loop throughout the entire process from data acquisition to decision output: (1) Sensor Module (Bottom Sensing Layer): Employs a multimodal sensor network with an IP68 protection rating (resistant to high temperatures up to 80°C and high salt spray corrosion, meeting environmental requirements). Specifically includes: 1) Strain gauge (installed on the critical section of the pile leg, sensitivity, sampling rate 10Hz): monitors local strain, taking into account the influence of prestress.

[0087] 2) Inclinometer (accuracy 0.01°, installed at the four corners of the deck): monitors overall tilt and deflection.

[0088] 3) GNSS: Real-time tracking of settlement and displacement, resisting multipath effects.

[0089] 4) Vibration sensor (triaxial, frequency response 0-100Hz): monitors vortex-induced vibration and seismic response, and simulates wave loads based on the Morison equation.

[0090] 5) Sensor redundancy design: at least two backups at each critical point, with total power consumption <200W (only core equipment operates during typhoon season). Installation scheme: strain gauges are welded / attached to the pile legs (location based on hotspot analysis), and the GNSS antenna is placed on top of the platform.

[0091] (2) Data Acquisition Module (Transmission Layer): Based on wireless sensor networks, with a sampling frequency of 10-50Hz, supporting 5G edge transmission (latency <100ms). Data preprocessing includes filtering (low-pass filter, cutoff frequency 5Hz) and compression (dimensionality reduction, reducing transmission volume by 50%). The platform integrates existing systems (such as the stress / tilt meter of Haiyang Shiyou 165) and embeds them with compatibility via the Modbus protocol to achieve data fusion.

[0092] (3) Data processing module (analysis layer): Deployed in the edge computing unit, running the core algorithm: 1) Confidence interval analysis submodule: Based on statistics, random loads are removed.

[0093] 2) Prestress Compensation Submodule: Finite Element Driven Stress Mapping.

[0094] 3) Fatigue prediction submodule: LSTM network, training dataset from historical monitoring (six months of data > points). Computational complexity and real-time performance <1 second.

[0095] (4) Early warning module (decision level): host computer software, threshold settings: settlement change alarm, stress fatigue life early warning. Outputs include visualization dashboard (Matplotlib charts), report generation and push, and support remote access.

[0096] It should be noted that, regarding integration, verification, and evaluation, this embodiment can achieve the following aspects: Integration: Embedded into the existing system of Haiyang Shiyou 165 (stress / tilt meter), data flow: sensor → wireless network → edge electronic device → cloud platform.

[0097] Verification: On-site deployment, six-month data evaluation: accuracy 98% (error <0.5mm), false alarms <3%. Risk analysis: equipment reliability (redundancy backup); baseline error (multi-point calibration). 3 years (20% reduction in maintenance).

[0098] The present invention can achieve the following technical effects: (1) Based on confidence interval analysis, the accuracy of stripping random loads is >95% (80% for the relevant technical threshold method). (2) Prestress compensation error <5%; (3) Early warning and prediction, fatigue assessment accuracy >95% (4) Low power consumption design, highly adaptable; (5) Compute edge-cloud coordination (can be optimized in the cloud).

[0099] See Figure 2 This embodiment is compared with various methods in related technologies.

[0100] like Figure 3 As shown, in other self-installing offshore oil platform structural safety monitoring methods proposed in this embodiment, the structural safety monitoring system includes a sensor network, a data acquisition module, a data processing module, an early warning module, and a digital twin module (cloud).

[0101] The sensor network is installed on a self-mounted offshore oil platform and is suitable for collecting structural response data of the platform. The sensor network includes various sensors, manufactured using high-strength, corrosion-resistant materials, and has a high protection rating. It is suitable for long-term stable operation in high-salinity, high-temperature, and high-pressure marine environments and is applicable to offshore platforms at various water depths.

[0102] The data acquisition module is located on the platform deck and connected to the sensor network. It receives, preprocesses, and stores monitoring data from the sensor network. Specifically, the module receives sensor data via wireless communication technology and integrates seamlessly with existing instruments on the platform, achieving efficient data fusion while keeping retrofit costs within budget. This makes it suitable for various marine engineering scenarios.

[0103] The data processing module, located on the platform deck or a remote server, is connected to the data acquisition module. This module generates a structural health status based on confidence interval analysis and prestress compensation processing of the monitoring data. Specifically, for example... Figure 1 As shown, data processing module 3 uses a high-performance computing unit, combined with finite element modeling (FEM), confidence interval (CI) analysis and machine learning (ML) algorithms, to evaluate the platform's settlement, irregular displacement, stress distribution, fatigue damage and vortex-induced vibration (VIV) in real time. The monitoring accuracy is high and conforms to international marine engineering standards (such as API and DNV standards).

[0104] The early warning module is located within the data processing module and connected to the remote control center. It generates early warning signals and pushes reports based on the processing results. Specifically, the early warning module uses confidence interval analysis to quickly trigger an early warning when monitored data deviates from the safe range. It pushes visual reports through various communication methods, supports on-site and remote decision-making, and has a low false alarm rate.

[0105] The digital twin module, located on a remote server and connected to the data processing module, is used to update the platform's digital model in real time and support predictive maintenance. Specifically, such as... Figure 4 As shown, the digital twin module dynamically updates the structural model based on finite element software and generates 3D visualization by combining real-time data, supporting predictive maintenance and cross-platform applications, and significantly reducing maintenance costs.

[0106] The structural safety monitoring system in this embodiment comprises a sensor network, a data acquisition module, a data processing module, an early warning module, and a digital twin module. The sensor network collects multi-dimensional structural response data in real time; the data acquisition module efficiently fuses and stores the data; the data processing module accurately assesses the structural state based on confidence interval analysis and prestress compensation; the early warning module predicts anomalies in advance; and the digital twin module supports dynamic modeling and predictive maintenance. Employing a low-power design, it is suitable for various marine environments (such as typhoons and deep seas), significantly reducing the risk of structural failure. It complies with international standards, achieving efficient, stable, and environmentally friendly intelligent operation and maintenance, and is applicable to marginal oil fields, deep-sea platforms, and other marine engineering structures.

[0107] Optionally, the sensor network includes strain gauges, inclinometers, positioning modules, and vibration sensors.

[0108] The strain gauges are installed at key locations on the platform's legs (made of high-strength steel with excellent mechanical properties). Multiple strain gauges are installed on each leg and fixed by welding or bonding to collect strain data and monitor areas of stress concentration. The strain gauges feature a high-sensitivity design, a high level of protection, and redundant configuration to ensure reliable data acquisition, making them suitable for high-salinity, high-pressure marine environments.

[0109] The inclinometer is installed at a key location on the platform deck to monitor the platform's deflection and tilt angle. It has high precision and a wide measurement range, is suitable for detecting irregular displacements, and can withstand harsh environments.

[0110] The positioning module is located at the top of the platform and uses high-precision positioning technology to monitor the platform's settlement (including static and dynamic displacement). It supports long-term monitoring, has high accuracy, and is suitable for various water depth conditions.

[0111] Vibration sensors are installed at key locations on the pile legs to monitor vortex-induced vibration and seismic response. They have a wide frequency response and high sensitivity, making them suitable for complex marine load environments.

[0112] The sensor network, through its low-power design and redundant configuration, ensures high data acquisition rates, is suitable for various marine environments (such as typhoon season and high salinity), and complies with relevant standards. Therefore, the sensor network can accurately acquire multi-dimensional structural response data, covering key indicators such as settlement, stress, and vibration, reducing the risk of missed detections, improving monitoring reliability, and is applicable to various marine platforms.

[0113] The data acquisition module includes a wireless transmission module, an edge computing unit, and a data storage module.

[0114] The wireless transmission module employs efficient wireless communication technologies (such as LoRa or ZigBee) to receive time-series data from sensor networks. It supports multi-node concurrency, long transmission distance, low latency, and uses encryption protocols to ensure data security. The module integrates with existing instruments on the platform through standard protocols (such as Modbus) to achieve data fusion. It has low modification costs and is suitable for various marine engineering scenarios.

[0115] The edge computing unit uses high-performance embedded computing devices to perform data preprocessing, including outlier removal and smoothing filtering.

[0116] Anomaly removal is based on statistical criteria to remove outlier data and reduce the impact of noise.

[0117] Smoothing filtering uses a sliding window method to smooth time-series data.

[0118] This method assumes that the noise is a zero-mean random process. After smoothing, it can preserve low-frequency trends (such as structural deformation) while filtering out high-frequency noise (such as sensor jitter or environmental interference). This significantly reduces noise and improves data quality.

[0119] The data storage module employs an embedded database to store long-term monitoring data, supporting local storage and cloud backup. Data is transmitted with encryption to ensure security and traceability. The data acquisition module reduces reliance on the cloud through edge computing, making it suitable for harsh marine environments and compliant with relevant regulations. Therefore, the data acquisition module ensures efficient data acquisition, preprocessing, and storage, supporting real-time monitoring and long-term data analysis, making it suitable for marginal oil fields and deep-sea platforms.

[0120] The data processing module includes a finite element modeling module, a confidence interval analysis module, and a machine learning module.

[0121] The finite element modeling module uses finite element software (such as ABAQUS or ANSYS) to create a 3D model of the platform's pile legs, including key geometric and material parameters (high-strength steel with excellent mechanical properties). The mesh uses high-precision shell elements, and boundary conditions simulate fixed and multi-source loads (self-weight, waves, wind loads, and corrosion) from the seabed. Wave loads are based on the Morison equation.

[0122] The confidence interval analysis module performs statistical analysis on the settlement data and calculates the confidence interval.

[0123] The machine learning module uses a Long Short-Term Memory (LSTM) network to predict fatigue damage.

[0124] The data processing module significantly improves monitoring accuracy by calibrating prestress with FEM, quantifying uncertainty through CI analysis, and predicting fatigue with LSTM. This is superior to traditional methods and complies with relevant standards.

[0125] The early warning module includes an anomaly detection module, a report generation module, and a communication module.

[0126] The anomaly detection module is based on confidence interval analysis, which compares the monitoring data with the benchmark interval in real time. When the data deviates from the safe range, an early warning is triggered. The response time is short and the false alarm rate is low.

[0127] The report generation module generates visual reports (settlement trend, stress distribution, fatigue prediction) using a graphical interface, supporting both on-site and remote viewing. Example: Anomalies in settlement are detected under extreme conditions (such as a typhoon), generating a report titled "Inspect Key Parts of Pile Legs".

[0128] The communication module pushes early warning signals through multiple communication methods, employs encryption protocols, has low latency, and supports real-time response from remote control centers. The early warning module achieves rapid early warning with a low false alarm rate, outperforming traditional threshold methods and meeting relevant regulatory requirements.

[0129] The digital twin module includes a model update unit and a visualization unit.

[0130] The model update unit dynamically updates the structural model based on finite element software, and combines real-time data (settlement, stress, vibration) to reflect structural changes under multi-source loads with low error.

[0131] The visualization unit generates 3D visualizations of stress fields, settlement trends, and fatigue damage, supporting predictive maintenance and reducing maintenance costs. The digital twin module enhances system scalability, making it suitable for various offshore platforms and compliant with relevant specifications.

[0132] Specifically, the execution process of this embodiment can be divided into the following stages: Deployment phase: Install sensor networks in key parts of the platform (such as legs, deck, and top), configure wireless transmission and edge computing units, and verify sensor functionality (high precision, low power consumption).

[0133] Specifically, this step aims to establish the physical foundation of the system, ensuring reliable deployment of sensor modules in various marine environments, such as high salinity, high pressure, and typhoon conditions. Prior to deployment, an on-site structural assessment is conducted, referencing environmental load specifications to identify key monitoring areas: the base of the pile legs (high-stress zone), the deck connection (deflection-sensitive area), and the platform top (settlement benchmark). The sensor network includes various sensors (such as strain gauges, inclinometers, positioning modules, and vibration sensors), manufactured using high-strength, corrosion-resistant materials, possessing a high protection rating, and suitable for long-term immersion in seawater. Installation methods include welding, bonding, or clamping, ensuring a tight fit between the sensors and the platform structure without affecting the platform's overall mechanical performance. The wireless transmission unit employs an encrypted protocol, supports multi-node concurrent data transmission, and boasts long transmission distances and low latency. The edge computing unit utilizes embedded devices with high-performance processing capabilities for initial data verification. The verification process includes functional testing (such as sensor response time and data consistency) and environmental adaptability checks (such as salt spray corrosion testing) to ensure the system possesses high accuracy and low power consumption characteristics immediately after deployment. The advantage of this step is that it reduces the impact of modifications on platform operation through modular deployment and redundant design, complies with structural design specifications, is applicable to various self-installation platform types (such as fixed or semi-fixed), and enables rapid deployment and low-risk integration.

[0134] Initialization phase: Establish a benchmark finite element model, calibrate the prestress, calculate the benchmark confidence interval, and train the machine learning model (using historical and synthetic data).

[0135] Specifically, this step is used to construct a benchmark reference for the system, ensuring the accuracy and comparability of subsequent monitoring data. The finite element modeling unit establishes a 3D model of the platform structure based on high-strength material parameters (such as the elastic modulus and Poisson's ratio of steel), simulating multi-source loads (such as self-weight, waves, wind loads, and corrosion effects). Fine mesh generation and boundary conditions are used to simulate the fixed state of the seabed, outputting the initial stress field and displacement distribution as a benchmark for prestress compensation. Prestress calibration establishes a stress mapping relationship by fitting sensor measurement data with model results. The calibration coefficients are optimized using the least squares method, with errors controlled within a safe range, suitable for various platform materials and structural forms. The confidence interval calculation unit performs statistical analysis on the initial settlement data, uses a sliding window for smoothing, and calculates the confidence interval. The benchmark interval is determined based on long-term stable operating conditions and is used for subsequent anomaly detection. Machine learning model training uses historical monitoring data and extreme operating condition data synthesized by a generative adversarial network. Inputs include stress amplitude, cycle count, and environmental factors; output is fatigue damage prediction. A long short-term memory network structure is used, with the loss function being the mean squared error. The training process includes data preprocessing, model optimization, and cross-validation to ensure generalization ability. The advantage of this step is that, through benchmarking and model training, the system becomes adaptive, reduces initial errors, conforms to fatigue design specifications, and is suitable for monitoring the entire lifecycle of marginal oilfield platforms from installation to operation, achieving efficient initialization and long-term reliability.

[0136] Operational monitoring phase: real-time data acquisition (high-frequency sampling), preprocessing (anomaly removal, smoothing filtering), calculation of confidence intervals, stress mapping, prediction of fatigue damage, and monitoring of vortex-induced vibration.

[0137] Specifically, this step is the core operational link of the system, ensuring continuous tracking of the platform's structural status. Real-time acquisition involves high-frequency sampling of multi-dimensional data (such as settlement, strain, vibration, and tilt angle) via a sensor network, and the data is transmitted to the edge computing unit. Preprocessing includes anomaly removal (eliminating noise points based on statistical criteria) and smoothing filtering (using a moving average window method to reduce high-frequency interference) to ensure high data quality. The confidence interval analysis unit calculates confidence intervals for the settlement data, compares the real-time interval with the baseline interval, and identifies anomalies if there is no overlap, removing the influence of random loads (such as dynamic responses to wind, waves, and currents). The stress mapping unit, based on a prestress compensation model, converts sensor measurement data into an absolute stress distribution, supporting multi-point calibration. The fatigue damage prediction unit uses a machine learning model, taking time-series data as input, and outputs damage assessments based on fatigue accumulation criteria, supporting early warning. The eddy-induced vibration monitoring unit captures response signals through vibration sensors, analyzes vibration frequency and amplitude using a fluid dynamics model, and assesses structural stability. The advantage of this step is that, through continuous monitoring and multi-algorithm collaboration, the system achieves high-precision status assessment, which is superior to the traditional threshold method (which has a high false alarm rate), complies with marine operation specifications, is applicable to platform operation and maintenance under various working conditions, and improves operational efficiency and safety.

[0138] Early warning phase: Anomalies are detected (deviation from the confidence interval), a visual report is generated, and an early warning signal is pushed out. Example: Anomalies in settlement trigger a "Check critical areas" alarm.

[0139] Specifically, this step is used to respond promptly to structural risks and ensure platform safety. The anomaly detection unit, based on confidence interval analysis, compares monitoring data with a baseline interval in real time. When data deviates from the safe range (e.g., settlement intervals do not overlap), an early warning signal is immediately triggered, supporting multi-level threshold settings (minor anomalies trigger a notification, severe anomalies trigger an emergency alarm). The visualization report generation unit uses graphical tools to display settlement trends, stress distribution, and fatigue prediction curves, including a timeline and anomaly markers, facilitating quick interpretation by operators. The push mechanism sends alarms through various communication methods (such as SMS or remote terminals), ensuring low latency and encrypted transmission for information security. Example: If an anomaly is detected under extreme conditions, the system generates a report and pushes a "Check critical areas" alarm, supporting on-site response. The advantages of this step are rapid early warning response, low false alarm rate, reduced manual intervention through visualization and push mechanisms, compliance with early warning requirements, applicability to high-risk scenarios such as typhoons or earthquakes, and enabling preventative maintenance and risk minimization.

[0140] Evaluation phase: Integrate long-term data to generate a comprehensive report (settlement trend, stress distribution, fatigue prediction) and verify system performance.

[0141] Specifically, this step is used for system performance verification and output, ensuring the traceability and usability of monitoring data. The data integration unit summarizes long-term monitoring data (such as settlement trends, stress distribution, and fatigue prediction), uses statistical tools to analyze the overall health status, and generates a comprehensive report, including multi-dimensional charts and risk assessments. System performance verification includes accuracy testing (comparison with benchmark data), stability checks (long-term operating rate), and compatibility assessment (integration with existing instruments). The report supports intellectual property applications, such as patent descriptions of technical features and software copyright registration of algorithm code. The advantage of this step is that, through data integration and verification, the system possesses self-assessment capabilities, complies with in-service assessment standards, is suitable for project acceptance and academic output, and improves technology transfer efficiency.

[0142] Expansion phase: Update the digital twin model and extend it to other marine engineering scenarios (such as deep-sea risers) to optimize maintenance strategies.

[0143] Specifically, this step is used for system expansion and optimization, supporting cross-platform applications. The digital twin model update unit dynamically adjusts the structural model based on real-time data, integrating finite element simulation and machine learning prediction to generate 3D visualizations and support virtual testing (such as load change simulation). Extension to other scenarios includes deep-sea riser monitoring (vortex-induced vibration analysis) and floating LNG unit structural assessment, achieving software reuse through modular design. Optimized maintenance strategies include damage-predictive scheduling algorithms to reduce on-site intervention. The advantages of this step are that the digital twin module enhances system flexibility, conforms to the digital twin framework, is applicable to various marine engineering projects (such as offshore wind power and deep-water drilling), and enables expansion from self-installed platforms to integrated operation and maintenance, improving economy and sustainability.

[0144] The structural safety monitoring system of this invention has the following beneficial effects: Highly efficient monitoring: Confidence interval analysis quantifies data uncertainty, prestress compensation calibrates stress, machine learning predicts fatigue damage, and monitoring accuracy is high, superior to the traditional threshold method (which has a higher false alarm rate).

[0145] High stability: The sensor network adopts a high protection level and redundant design, which is suitable for harsh marine environments, has a high data acquisition rate, supports long-term stable operation, and meets relevant specifications.

[0146] Environmental protection: Low power consumption design reduces energy consumption, accurate monitoring reduces maintenance frequency, minimizes disturbance to the marine ecosystem, and complies with relevant regulations.

[0147] Easy to operate: The embedded system integrates seamlessly with existing instruments, the graphical interface supports remote operation, automatic early warning reduces manual intervention, and meets relevant specifications.

[0148] Operational reliability: Multi-algorithm collaboration (finite element method + confidence interval + machine learning) and edge computing ensure stable operation of the system in complex environments, meeting relevant specifications.

[0149] Highly scalable: The digital twin module supports dynamic modeling and predictive maintenance, reducing maintenance costs, and is suitable for various marine engineering scenarios, meeting relevant regulatory requirements.

[0150] like Figure 5 As shown in the figure, this embodiment proposes a self-installing offshore oil platform structural safety monitoring device, including: Acquisition unit 501 is used to acquire the pre-processed strain time series data of the legs of the self-installed offshore oil platform during the first time period. Analysis unit 502 is used to perform stationarity analysis based on confidence intervals on preprocessed strain time series data to obtain analysis results; The stripping unit 503 is used to identify and strip the random load components in the preprocessed strain time series data according to the analysis results, so as to obtain the target strain time series data. The mapping unit 504 is used to perform finite element-driven prestress compensation and stress mapping on the target strain time series data to obtain key stress time series data. Prediction unit 505 is used to predict the cumulative fatigue damage index of the pile leg in the second time period based on key stress time series data; wherein the second time period is after the first time period. Correction unit 506 is used to perform corrosion coupling correction on the cumulative fatigue damage index to obtain the target damage index; The early warning unit 507 is used to determine whether to issue a fatigue warning based on the target damage index.

[0151] It should be noted that the processing procedures of the acquisition unit 501, analysis unit 502, stripping unit 503, mapping unit 504, prediction unit 505, correction unit 506, and early warning unit 507, as well as their beneficial effects, can be referred to respectively. Figure 1 Steps S101 to S107 are not described in detail here.

[0152] Optionally, the preprocessed strain time series data includes preprocessed strain values ​​corresponding to multiple consecutive timestamps; Analysis unit 502 is also used for: Calculate the mean and standard deviation using the preprocessed strain value corresponding to each timestamp; The preprocessed strain value, mean, and standard deviation corresponding to each time stamp are input into the created confidence interval calculation model to determine the confidence interval corresponding to each time stamp; For any timestamp, determine whether there is an intersection between the confidence interval corresponding to the timestamp and the confidence interval corresponding to the previous timestamp; If there is an intersection, it is determined that the stationarity of the preprocessed strain value corresponding to the timestamp meets the requirements; If there is no intersection, it is determined that the stationarity of the preprocessed strain value corresponding to the timestamp does not meet the requirements.

[0153] Optionally, analysis unit 502 is also used for: Determine the first upper bound and the first lower bound in the confidence interval corresponding to the timestamp, and determine the second upper bound and the second lower bound in the confidence interval corresponding to the previous timestamp; Determine the minimum value between the first upper bound and the second upper bound, and determine the maximum value between the first lower bound and the second lower bound; Subtract the maximum value from the minimum value to obtain the corresponding difference; Select the maximum value between 0 and the difference and use it as the intersection measure; If the intersection measure is not equal to 0, then it is determined that the confidence interval corresponding to the timestamp and the confidence interval corresponding to the previous timestamp have an intersection. If the intersection measure is equal to 0, it is determined that the confidence interval corresponding to the timestamp and the confidence interval corresponding to the previous timestamp do not intersect.

[0154] Optionally, the stripping unit 503 is also used for: Each pre-processed strain value in the pre-processed strain time series data that does not meet the stationarity requirement is identified as a random load component. Each random load component in the preprocessed strain time series data is deleted to obtain the target strain time series data.

[0155] Optionally, mapping unit 504 is also used for: A three-dimensional model of the pile leg is constructed based on the finite element method and the set model conditions. The self-weight stress of the pile leg is simulated using the three-dimensional model to obtain the self-weight stress distribution of the pile leg. The target strain time series data and the stress distribution of the pile leg self-weight are input into the created mapping model to perform prestress compensation and stress mapping, and key stress time series data are obtained. The mapping model is as follows: ; in, For key stress time series data, To determine through static simulation using finite element modeling, For target strain time series data, The stress distribution due to the self-weight of the pile leg. For corrosion reduction factor, To obtain corrosion stress time series data.

[0156] Optionally, prediction unit 505 is also used for: Key stress time series data are input into the trained time series prediction model so that the time series prediction model can predict the cumulative fatigue damage index of the pile leg in the second time period. The correction unit 506 is also used for: The corresponding corrosion coupling index is determined based on the cumulative fatigue damage index. The cumulative fatigue damage index is added to the corrosion coupling index to obtain the corresponding target damage index; The decision unit is also used for: Determine whether the target damage index is greater than a set threshold; If the target damage index is greater than the set threshold, a fatigue warning will be issued; If the target damage index is not greater than the set threshold, fatigue warning is prohibited.

[0157] Optionally, the acquisition unit 501 is also used for: Obtain the original strain time series data of the pile leg during the first time period; Outlier identification and removal are performed on the original strain time series data to obtain strain time series data after removal. Based on the moving average window, the strain time series data after rejection is smoothed and filtered to obtain the filtered strain time series data, which is used as the preprocessed strain time series data.

[0158] The self-installing offshore oil platform structural safety monitoring device proposed in this embodiment can acquire pre-processed strain time-series data of the legs in the self-installing offshore oil platform during the first time period. It performs stationarity analysis based on confidence intervals on the pre-processed strain time-series data to obtain the analysis results, and identifies and removes random load components from the data based on the analysis results to obtain target strain time-series data. Finite element-driven prestress compensation and stress mapping are then performed on the target strain time-series data to obtain key stress time-series data. The cumulative fatigue damage index of the legs during the second time period is predicted based on the key stress time-series data; the second time period is after the first time period. Corrosion coupling correction is applied to the cumulative fatigue damage index to obtain the target damage index, and a fatigue warning is determined based on the target damage index. This invention achieves integrated confidence interval analysis, prestress compensation mapping, and fatigue damage prediction for safety monitoring, thereby realizing high-precision, low-power real-time monitoring and enhancing the reliability of structural safety monitoring for self-installing offshore oil platforms.

[0159] In this embodiment, the self-installing offshore oil platform structure safety monitoring device is presented in the form of a functional unit. Here, a unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0160] This invention also provides a computer device having the above-described features. Figure 5 The self-installing offshore oil platform structural safety monitoring device shown is shown.

[0161] Please see Figure 6 The present invention provides a schematic diagram of the structure of a computer device according to an optional embodiment. The computer device includes one or more processors 10, a memory 20, and interfaces for connecting the various components, including high-speed interfaces and low-speed interfaces. The various components are interconnected via different buses and can be mounted on a common motherboard or otherwise installed as needed. The processors can process instructions executed within the computer device, including instructions stored in or on memory to display graphical information of a GUI on an external input / output device (such as a display device coupled to the interface). In some optional embodiments, multiple processors and / or multiple buses can be used with multiple memories, if desired. Similarly, multiple computer devices can be connected, each providing some of the necessary operations (e.g., as a server array, a group of blade servers, or a multiprocessor system). Figure 6 Take a processor 10 as an example.

[0162] Processor 10 may be a central processing unit, a network processor, or a combination thereof. Processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CAMP), a field-programmable gate array (FPGA), a general-purpose array logic (GDA), or any combination thereof.

[0163] The memory 20 stores instructions executable by at least one processor 10 to cause at least one processor 10 to perform the method shown in the above embodiments.

[0164] The memory 20 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function. The data storage area may store data created based on the use of the computer device. Furthermore, the memory 20 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some alternative embodiments, the memory 20 may optionally include memory remotely located relative to the processor 10, which can be connected to the computer device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0165] Memory 20 may include volatile memory, such as random access memory. Memory may also include non-volatile memory, such as flash memory, hard disk, or solid-state drive. Memory 20 may also include combinations of the above types of memory.

[0166] The computer device also includes a communication interface 30 for communicating with other devices or communication networks.

[0167] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code, which, when accessed and executed by the computer, processor, or hardware, implements the methods shown in the above embodiments.

[0168] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for monitoring the structural safety of a self-installing offshore oil platform, characterized in that, include: Acquire pre-processed strain time-series data of the legs of the installed offshore oil platform during the first time period; A stationarity analysis based on confidence intervals is performed on the preprocessed strain time series data to obtain the analysis results. Based on the analysis results, the random load components in the preprocessed strain time series data are identified and removed to obtain the target strain time series data. Finite element-driven prestress compensation and stress mapping are performed on the target strain time series data to obtain key stress time series data; The cumulative fatigue damage index of the pile leg in the second time period is predicted based on the key stress time series data; wherein, the second time period is after the first time period; The cumulative fatigue damage index is subjected to corrosion coupling correction to obtain the target damage index, and fatigue warning is determined based on the target damage index.

2. The method according to claim 1, characterized in that, The preprocessed strain time series data includes preprocessed strain values ​​corresponding to multiple consecutive timestamps; The step of performing stationarity analysis based on confidence intervals on the preprocessed strain time series data to obtain analysis results includes: Calculate the mean and standard deviation using the preprocessed strain value corresponding to each timestamp; The preprocessed strain value, mean, and standard deviation corresponding to each timestamp are input into the created confidence interval calculation model to determine the confidence interval corresponding to each timestamp; For any of the timestamps, determine whether there is an intersection between the confidence interval corresponding to the timestamp and the confidence interval corresponding to the previous timestamp; If there is an intersection, it is determined that the stability of the preprocessed strain value corresponding to the timestamp meets the requirements; If there is no intersection, it is determined that the stability of the preprocessed strain value corresponding to the timestamp does not meet the requirements.

3. The method according to claim 2, characterized in that, The step of determining whether the confidence interval corresponding to the timestamp and the confidence interval corresponding to the previous timestamp intersect includes: A first upper bound and a first lower bound are determined in the confidence interval corresponding to the timestamp, and a second upper bound and a second lower bound are determined in the confidence interval corresponding to the previous timestamp. Determine the minimum value between the first upper bound and the second upper bound, and determine the maximum value between the first lower bound and the second lower bound; Subtract the maximum value from the minimum value to obtain the corresponding difference; The maximum value between 0 and the difference is selected as the intersection measure; If the intersection measure is not equal to 0, then it is determined that the confidence interval corresponding to the timestamp and the confidence interval corresponding to the previous timestamp have an intersection. If the intersection measure is equal to 0, then it is determined that the confidence interval corresponding to the timestamp and the confidence interval corresponding to the previous timestamp do not intersect.

4. The method according to claim 2, characterized in that, The step of identifying and removing random load components from the preprocessed strain time series data based on the analysis results to obtain the target strain time series data includes: Each pre-processed strain value whose stability requirement is not met in the pre-processed strain time series data is determined as a random load component. Each of the random load components in the preprocessed strain time series data is deleted to obtain the target strain time series data.

5. The method according to claim 1, characterized in that, The process of performing finite element-driven prestress compensation and stress mapping on the target strain time series data to obtain key stress time series data includes: A three-dimensional model of the pile leg is constructed based on the finite element method and the set model conditions. The self-weight stress of the pile leg is simulated using the three-dimensional model to obtain the self-weight stress distribution of the pile leg. The target strain time series data and the stress distribution of the pile leg self-weight are input into the created mapping model to perform prestress compensation and stress mapping, thereby obtaining the key stress time series data. The mapping model is as follows: ； in, The key stress time series data, To determine through static simulation using finite element modeling, For target strain time series data, The stress distribution of the self-weight of the pile leg is shown. For corrosion reduction factor, To obtain corrosion stress time series data.

6. The method according to claim 1, characterized in that, The step of predicting the cumulative fatigue damage index of the pile leg in the second time period based on the key stress time series data includes: The key stress time series data are input into the trained time series prediction model so that the time series prediction model can predict the cumulative fatigue damage index of the pile leg in the second time period. The step of performing corrosion coupling correction on the cumulative fatigue damage index to obtain the target damage index includes: The corresponding corrosion coupling index is determined based on the cumulative fatigue damage index. The cumulative fatigue damage index is added to the corrosion coupling index to obtain the corresponding target damage index; The step of determining whether to issue a fatigue warning based on the target damage index includes: Determine whether the target damage index is greater than a set threshold; If the target damage index is greater than the set threshold, a fatigue warning will be issued; If the target damage index is not greater than the set threshold, fatigue warning is prohibited.

7. The method according to any one of claims 1 to 6, characterized in that, The pre-processed strain time-series data obtained from the legs of the offshore oil platform during the first time period includes: Obtain the original strain time series data of the pile leg during the first time period; The original strain time series data is subjected to outlier identification and removal to obtain strain time series data after removal; Based on the moving average window, the strain time series data after removal is smoothed and filtered to obtain the filtered strain time series data, which is used as the preprocessed strain time series data.

8. A self-installing offshore oil platform structural safety monitoring device, characterized in that, include: The acquisition unit is used to acquire the pre-processed strain time series data of the legs of the self-installed offshore oil platform during the first time period. The analysis unit is used to perform stationarity analysis based on confidence intervals on the preprocessed strain time series data to obtain analysis results; The stripping unit is used to identify and strip the random load components in the preprocessed strain time series data according to the analysis results, so as to obtain the target strain time series data. The mapping unit is used to perform finite element-driven prestress compensation and stress mapping on the target strain time series data to obtain key stress time series data. The prediction unit is used to predict the cumulative fatigue damage index of the pile leg in a second time period based on the key stress time series data; wherein the second time period is after the first time period; A correction unit is used to perform corrosion coupling correction on the cumulative fatigue damage index to obtain the target damage index; The early warning unit is used to determine whether to issue a fatigue warning based on the target damage index.

9. A computer device, characterized in that, include: A memory and a processor are interconnected, the memory storing computer instructions, and the processor executing the computer instructions to perform the self-installing offshore oil platform structural safety monitoring method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to execute the self-installing offshore oil platform structural safety monitoring method according to any one of claims 1 to 7.

Images